Support vector machines always show a margin whose width changes. How can I find what my margin is ? Eg, if my weights are $[-1,1]$ what is my margin ? Also, do larger weights mean larger margin ?
1
-
1$\begingroup$ Have you seen the general discussion of SVMs at stats.stackexchange.com/questions/23391/…? Perhaps that will help. Otherwise, search our site: stats.stackexchange.com/search?q=%5Bsvm%5D+margin. $\endgroup$– whuber ♦Jan 4, 2018 at 14:05
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
Larger weights mean smaller margin. Given the weights you can compute the margin by 1/||w||.
-
$\begingroup$ thank you whuber and saibot. It was a matter of plugging in some values and vizualization. $\endgroup$– KongJan 4, 2018 at 14:37